sufficiently perfect to enable me to decide. This spe- 
cies of Centunculus was originally found in Nepaul, 
its rediscovery on the Neilgherries adds another to 
the many already existing links which connect these 
distant floras. 
2000. PRIMULA DENTICULATA P (Smith, Ex. Bot.), 
leaves rugous, thin, glabrous, ovato-lanceolate, un- 
equally denticulate, acutish, smooth above, beneath 
more or less dusted with white farina, sometimes 
without farina, narrowed into a winged petiol, sheath- 
ing and membranous at the base: involucrum many- 
flowered, leaflets acuminate, the exterior ones broader 
lanceolate longer than the pedicels, the interior ones 
linear lanceolate, shorter: calyx urseolate, 5-cleft, 
divisions linear lanceolate acute, longer than the 
tube: corolla salver-shaped, lobes obcordate, obtuse. 
Duby in D. C. Prod. 
The specimens from whieh the drawing was made 
were kindly communicated by Captain Munro who 
gathered them at Hungarung on the Himalayas in 
August. I am doubtful whether I have correctly 
named the plant as the specimens differ in some 
minute particulars from the character, but as they 
agree in their more prominent features, I could not 
venture on constituting this a new species, the more 
so as I have not an opportunity of consulting Smith's 
figure. Should it prove new, 1 would suggest its 
being dedicated to the discoverer. It is introduced 
here mainly to fill the plate, but also in the hope that, 
since Centunculus has been found on our southern 
mountains, a Primula may be found to bear it com- 
pany. 
# SCITAMINER. 
This is the Linnean name of a curious, beautiful and 
useful group of plants, including the Plantain, Car- 
damom, Ginger, Turmerie, Zedoary, Arrowroot, In- 
dian shot (canna), and many others. The original 
group, which now includes about 300 known species, 
is divided into three orders—Musacee, Zingiberacee, 
and Marantaceæ. The first is distinguished by hav- 
ing several stamens—the second by having one sta- 
men with a perfect 2-celled anther, and the 3d by 
having 1 stamen with a 1-celled or half anther, and 
that placed at one side of the flower, usually on a 
petaloid filament. 
Illustrations of each of these orders will be found 
among the following series of plates. The group, 
viewed as a whole, is readily distinguished by its 
foliage and habit. Erect herbaceous stalks with 
sheathing more or less lanceolate leaves, having a 
distinct mid rib with the lateral veins diverging 
thence at more or less acute angles towards the mar- 
gin. In most other monocotyledonous plants (some 
exceptions will be mentioned by and by) they run 
in parallel lines from the base to the apex. 
At first sight the flowers of Zingiberaceæ seem, as 
in orchids, to consist of a six-lobed perianth, 3 ex- 
terior and 3 interior, one of the latter more or less 
differing from the others, forming, as in orchids, a lip 
opposite the stamen. This is not, however, their true 
structure, forthey have in addition to this coroloid 
perianth, a distinet calyx, Sert f much shorter and 
embracing its tubular base) which is wanting in 
orchids. The difference is explained by assuming 
that in this group there are six stamens, 5 of which 
are modified, and only three in orchids, two of which 
are modified or wanting. That such is really the 
case is shown by the Plantain which has a six-parted 
perianth and six stamens one only of the latter im- 
perfect. In it, the 3 larger exterior lobes of the 
perianth correspond to the calyx of the Gingers, the 
3 smaller interior to the exterior lobes of the per- 
ianth, while the 3 outer stamens represent the inner 
perianth and the 3 inner the staminal series of Gin- 
gers. According to this view, the flower of Zingibe- 
raceæ consists, 1st of the calyx, 2d of the exterior or 
calycine lobes of the perianth, 3d, of the interior, or 
petaloid lobes, modified stamens, and 4th, of the pro- 
per stamens, two of which are abortive, and the third, 
or odd one, placed opposite the lip, perfect. Maran- 
tace? differ from this arrangement in perfecting one 
of the lateral stamens in place of the odd or pos- 
terior one. 
In the discrimination of the genera of Zingi- 
beraceæ the anther is usually looked to as furnish 
ing the essential characters, but of course the other 
parts of the flower are not overlooked. The first 
point to be noticed in examining one of these plants 
is to ascertain whether or not the filament extends 
beyond the anther. If it has not a crest or prolon- 
gation it is referable to Hedychium, Alpinia, Globba, 
Roscoea or Gastrochilus, all of which have crestless 
anthers, but are easily distinguished by other marks. 
The lateral appendages of the anthers of some of the 
Globbas do not come under that denomination. 
Of those that have it prolonged, Zingiber has an 
awl-shaped point. Elatteria, a short more or less 
fleshy thickening of the point. Costus, a short mem- 
branous prolongation. Curcuma, a dilated point and 
two spurs at the base. Kæmpferia, a long membra- 
nous forked point. Monolophus, a short broad re- 
flexed point. Roscoea has the base of the anther 
prolonged. 
These brief indications of the essential characters 
of the genera, so far as they are derived from the 
anther, will suffice to show that it is not generally 
difficult to distinguish the genera of this order with 
fresh plants in hand; and that, even with dried spes 
cimens if the flowers are not much injured in the 
drying, but a moderate degree of skill is required to 
open, for examination, flowers previously softened by 
immersion for a few minutes in hot water. 
The genera of Marantacee are easily known by 
their habit. 
Asitis probable my figures will generally be ex- 
amined in comparison with fresh plants with which 
in minute particulars they may not at all times be 
found to correspond, it is proper to mention that sev- 
eral of them are taken from dried specimens, and 
that in such cases minute accuracy of outline is not 
always attainable, even while the likeness is so well 
preserved as to leave no doubt as to the identity of 
the object represented. This remark is more espe- 
cially applicable to the magnified dissected flowers 
which, it must be allowed, it is often difficult so far 
to restore as to admit of the representation conveying 
a correct idea of the aspect of the parts, as seen in 
the growing plant, but Itrust that are generally so 
well done as to leave no doubt of the species to 
which they refer and which they are intended to 
make known. 
2001. GLOBBA MARANTINOIDES (R. W., G. maran- 
tina, R. W. Icon., non Willd.), leaves petioled, lance- 
olate : spike terminal, distichous, lower bracts bulbife- 
rous; upper floriferous: flowers 2-4 in each bract: 
lip entire, truncated at the apex, reflexed. 
L5 
